Turn supplied business evidence into a decision
Use only information supplied in the conversation and this skill text. Do not browse, call tools, inspect files, execute code, create artifacts, or take external actions. Return the work directly in chat. Attribute material claims to supplied source labels or quotations; a pasted URL is a source label, not evidence that its contents were checked. Distinguish supplied facts, reasonable interpretations, proposals, and unknowns.
Inputs and scope: Establish the decision owner, options, objective, horizon, constraints, and supplied financial or operating evidence. Identify which baseline represents the actual alternative to acting. If a decision-changing input is absent, ask the smallest useful question and complete the portions supported by available material. State assumptions explicitly; do not manufacture facts, approvals, dates, or completion evidence.
Method 1. Write the decision as a choice with a consequence and deadline when supplied. Separate the business outcome from a preferred solution; retain a no-change option when it is viable. 2. Create a compact evidence ledger containing claim, source, period, population, measure, and limitation. Reconcile conflicting definitions before combining figures and distinguish reported measurements from stakeholder estimates. 3. Normalize comparisons using compatible units, currency, time horizon, customer cohorts, and scope. Show only bounded arithmetic supported by supplied numbers; explain exclusions and avoid comparing revenue with contribution or recurring with one-off costs. 4. Assess opportunity magnitude and addressability separately. Identify the total observed problem, the portion within the organization's control, adoption or capacity constraints, and the mechanism by which an option could improve the result. 5. Compare alternatives against decision-relevant criteria such as impact, cost, reversibility, operational burden, timing, and downside. Use qualitative judgments when denominators or probabilities are absent rather than inventing precise scores. 6. Recommend a choice or a specific information-gathering decision. State the assumptions that could reverse it, an observable success criterion, and a sensible review point. Treat modeled benefits as hypotheses, not realized value.
Output: Return the decision, evidence summary, normalized option comparison, recommendation, assumptions, and the next decision-changing question or test.
Quality checks: Confirm the baseline, horizon, units, arithmetic, controllable portion, and source support. Avoid double-counting benefits that describe the same underlying improvement; identify material omitted costs.
Worked example: A team handles 200 requests monthly, spending 15 minutes each. A proposed template could address 80 requests and is estimated to save 5 minutes each. The addressable estimate is 400 minutes monthly, not the full 50 hours currently spent. Recommend a trial if its setup burden is acceptable, with measured handling time and rework as the review evidence.